the level of survival has increased for both Blacks and Whites, but the increase has been more rapid for Whites than Blacks; thus, the racial disparity has widened. White males, during 1989 to 1994, had a 60 percent survival rate compared to 45 percent of Black males. There is some variation by specific type of cancer, with the racial disparity being greater for colon and prostate cancer than for lung cancer.

The racial difference in cancer survival for females is similar to that observed for males. During 1974 to 1979, there was a 10-point difference in cancer survival, which increased to a 14-point difference during 1989 to 1994. During the latter period, Black females had a 49 percent, five-year survival rate compared to 63 percent for White females. This pattern is similar to that of males, both in the magnitude of the difference and the widening gap in survival rates over time. Also, mirroring the pattern observed for males, Black females have a very small survival rate difference for lung cancer, with much larger differences for colon and breast cancer. The increasing racial disparity in survival is especially striking for colon cancer, with the Black-White difference being almost three times as large in the early 1990s as it was in the 1970s. The racial difference in survival rate for breast cancer is also instructive. Breast cancer stands out as one of the few physical health outcomes for which the incidence rate is higher for Whites than for Blacks. Yet in spite of the fact that Black females are less likely to get breast cancer, they are more likely to die from it. It is well documented that there are racial differences in cancer staging, with Black females more likely to have more advanced cancer at the time of detection than White females. Some research suggests that Black females have even poorer stage-specific survival rates than their White counterparts (Hunter et al., 1993).

A third factor that needs to be kept in mind, to put these data into perspective, is the dramatic heterogeneity that exists within each of the major racial categories. Subgroup variations within the major racial categories tend to predict variation in sociodemographic and socioeconomic characteristics in access to and use of medical care and in health status.

For example, the Native American or Alaska Native category consists of more than 250 federally recognized tribes; 209 Alaska Native villages; 65 communities that have been recognized as tribes by the states in which they are located, but not by the federal government; and several dozen other communities that have not received any formal recognition (Norton and Manson, 1996). Although these tribes and communities share a common history of exploitation and oppression, there is great diversity in cultures, socioeconomic circumstances, and health. The Indian Health